flavour physics after the first run of the lhc
Post on 24-Feb-2016
32 Views
Preview:
DESCRIPTION
TRANSCRIPT
1
Flavour physics after the first run of the LHC
Diego Martinez Santos (NIKHEF, VU Amsterdam and CERN)
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze,
10/07/2013
On behalf of theLHCb Collaboration
Including also ATLAS and CMS results
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
2
Introduction
The SM is very successful in accurately describing accelerator data, yet it is known to be an incomplete theory (gravity, DM, etc…)
Several alternatives/extension exist (SUSY, compositeness, ED’s … ) which could be a better approach to nature
Need (new) experimental data to:Break SM (if possible) at acceleratorsConstrain BSM parameters, rule out models inconsistent with
data …
Two main approaches at LHCDirect search of new particles ATLAS/CMSIndirect search LHCb
• Access to BSM physics through its effect in B,D,K,τ decays• This approach has been very successful in many cases:
top quark or Z0 were inferred from indirect effects many years before direct observation
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
3
A nice parallelism ?
The Ptolemaic model was very successful on precisely describe all astronomic data for many years. (very much like SM)
Alternate (i.e. heliocentric) models existed since Aristarchos (c.III BC), but they predicted unobserved phenomena like parallax -not observed till c.XIX-, which could only fit if one puts the distance of the stars at a very large scale. (very much like BSM)
In c.XVII, Galileo points the first telescope to the sky and observes a series of phenomena that contradicted Ptolemaic model, and favoured heliocentric theories…. (very much like LHC ??)
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
4
Flavour physics results from LHCb
• Very rare decays
• Light flavoured mesons decaying into dimuons:• Bs → µµ, Bd → µµ, D0 → µµ, KS → µµ
• Other very rare decays
• Results in CPV
• Electroweak phase • First observation of CPV in Bs decays• CPV in charm
• The rare decay Bd → K* µµ
Nice flavour topics not covered here: radiative decays, measurement of CKM angle γ
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
5
Flavour physics results from LHCb
• Very rare decays
• Light flavoured mesons decaying into dimuons:• Bs → µµ, Bd → µµ, D0 → µµ, KS → µµ
• Other very rare decays
• Results in CPV
• Electroweak phase • First observation of CPV in Bs decays• CPV in charm
• The rare decay Bd → K* µµ
I will also cover some ATLAS and CMS results
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
6
VERY RARE DECAYS
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
7
Bs(d) → µµThese decays are very supressed in SM
BR(Bs → µµ) = (3.54 ± 0.30)x10-9(time averaged)
arXiv:1208.0934.BR(Bd → µµ) = (1.07 ± 0.10)x10-10
Scenario Would point to
BR(Bs→µµ) >> SMBig enhancement from NP in the scalar sector, SUSY at high tanβ
BR(Bs→µµ) ≠ SM SUSY, ED’s, LHT, TC2
BR(Bs→µµ) ≈SM Anything ( rule out regions of parameters space that predict sizable departures w.r.t SM)
BR(Bs→µµ) <<SM NP in the scalar sector, but full MSSM ruled out. NMSSM (Higgs singlet) good candidate
BR(Bs→µµ) /BR(Bd→µµ) )≠ SMCMFV ruled out. New FCNC independent of CKM matrix (RPV-SUSY, ED’s ,etc …)
… but can be modified by NP. Here you have a rough table of what would imply each potential result (note tha the arrow goes only in one direction)
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
8
Bs(d) → µµ (LHCb analysis strategy) PhysRevLett.110.021801arXiv:1211.2674
I) Selection cuts in order to reduce the amount of data to analyze.
II) Classification of Bs,d→μμ events in bins of a 2D space
- Invariant mass of the μμ pair - Boosted Decision Tree (BDT) combining geometrical and kinematical information about the event.
III) Control channels (B→hh, B → J/ψK, mass sideb.) to get signal and background expectations w/o relying on simulationIV) Use CLs,b for limits and signal significance. Also fit for signal strength
arXiv:1211.2674
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
9
Bs(d) → µµ (LHCb results) PhysRevLett.110.021801arXiv:1211.2674
BR(Bs → µµ) fit:
BR(Bd → µµ) CLs limits:
1 fb-1 from 2011 (7 TeV) and 1 fb-1 from 2012 (8 TeV) are statistically combined.
We see a 3.5σ signal in the Bs mode
No significant (~1.3σ) signal (yet) in the Bd
Update expected (very) soon !
arXiv:1211.2674
arXiv:1211.2674
10BSM after the first run of the LHC. Galileo Galilei Institute. Firenze,
2013
Bs(d) → µµ (ATLAS / CMS/averages)
Both experiments perform cut-based analyses (MVA under development, afaik). Up to now both show less sensitivity than LHCb for the same integrated luminosity (due to trigger/reconstruction/resolution).
However, for the same time period ,CMS has been performing almost equally well than LHCb.
Experiment
Luminosity (fb-1) 2.4 5.0
95%CLs (Bs) (10-
9)22 7.7
95%CLs (Bd) (10-
9)- 1.8
arXiv:1204.0735 arXiv:1203.3976
Latest LHC combination is ~old (only 0.4 fb-1 from LHCb)Mastercode’s private/unofficial combination yields
BR(Bs→μμ)
http://mastercode.web.cern.ch/mastercode/news.php
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
11
Bs(d) → µµ (what does it imply?)
Scenario Would point to
BR(Bs→µµ) >> SMBig enhancement from NP in the scalar sector, SUSY at high tanβ
BR(Bs→µµ) ≠ SM SUSY, ED’s, LHT, TC2
BR(Bs→µµ) ≈SM Anything ( rule out regions of parameters space that predict sizable departures w.r.t SM)
BR(Bs→µµ) <<SM NP in the scalar sector, but full MSSM ruled out. NMSSM (Higgs singlet) good candidate
BR(Bs→µµ) /BR(Bd→µµ) )≠ SMCMFV ruled out. New FCNC fully independent of CKM matrix (RPV-SUSY, ED’s ,etc …)
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
12
Bs(d) → µµ (what does it imply?)
Scenario Would point to
BR(Bs→µµ) >> SMBig enhancement from NP in the scalar sector, SUSY at high tanβ
BR(Bs→µµ) ≠ SM SUSY, ED’s, LHT, TC2
BR(Bs→µµ) ≈SM Anything ( rule out regions of parameters space that predict sizable departures w.r.t SM)
BR(Bs→µµ) <<SM NP in the scalar sector, but full MSSM ruled out. NMSSM (Higgs singlet) good candidate
BR(Bs→µµ) /BR(Bd→µµ) )≠ SMCMFV ruled out. New FCNC fully independent of CKM matrix (RPV-SUSY, ED’s ,etc …)
… You expect some constraints at least in SUSY
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
13
Bs(d) → µµ (what does it imply?)
NUHM1 global fit 1fb-1 LHC . arXiv:1207.7315
(Expectations as for 2011)
arXiv:1112.3564
Constraints are model dependent, but the usual tendency is that Bs → µµ dominates at high tanβ/MALikelihoods of global fits get modified. No big effect expected in the p-value
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
14
Bs(d) → µµ (what does it imply?)
Constraints are model dependent, but the usual tendency is that Bs → µµ dominates at high tanβ/MALikelihoods of global fits get modified. No big effect expected in the p-value A FAQ: How big is the SUSY phase space ruled out by Bs → µµ? Outside the effect on the likelihoods, the question has no objective answer:
• All values of the fundamental parameters equiprobable? the excluded volume is O(5%). But this is not invariant under reparameterization
• All the previously allowed BR’s equiprovable? the excluded area is large. But also no reason to consider all BR’s are equiprobable a priori (same as above)
• It’s a bit like asking which is the fraction of the SM parameter space that has been ruled out up to know by current data
NUHM1 global fit 1fb-1 LHC . arXiv:1207.7315
(Expectations as for 2011)
arXiv:1112.3564
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
15
KS → µµ, D0 → µµ
arXiv:1209.4029
Bs → µµ
Bd → µµ
D0→µµ KL → µµ
KS → µµ
3.2 <0.94 <7.6 6.84±0.11
11
LHCb LHCb LHCb BNL E871 LHCb
BR of neutral flavoured mesons into dimuons [x10-9]
R-S warped EDarXiv:1212.4849
LHCb also sets world best upper limit in other dimuon decays
arXiv:1305.5059
SM BR(Ks →µµ) =(5.1±1.5)x10-12
SM prediction: BR(D0→µµ) < 1.6x10-11
(depends on knowledge of BR(D0→γγ) )
arXiv:hep-ph/0311084
Ks →µµ
BR(KS →µµ) is sensitive to different physics than BR(KL →µµ). Limits at the 10-
11, 10-12 quite interesting specially if NP is found at NA62
Limits at the 10-8 – 10-9 level
BR(D0→µµ) at the 10-10 level can be sensitive to ED and RPV.LHCb will explore those ranges with the upgraded detector.
arXiv:1209.4029
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
16
Other very rare decays @ LHCb
Decay Main BSM test 95% upper limitBs→µµµµ Some SUSY
scenarios<1.6x10-8 (arXiv:1303.1092)
Bd→µµµµ Some SUSY scenarios
<6.6x10-9 (arXiv:1303.1092)
τ→µµµ LFV (ex: LHT) <8.0x10-8 (arXiv:1304.4518)(still below B-factories sensitivity)
τ→pµµ LNV, BNV (“”) <4.4x10-7 proton<3.3x10-7 anti-proton(arXiv:1304.4518)
Bs→eµ RPV, Pati-Salam LQ…
<1.4x10-8 (LHCb-PAPER-2013-030)
Bd→eµ RPV, Pati-Salam LQ…
<3.7x10-9 (LHCb-PAPER-2013-030)
B→Xµ+µ+ 4th gen. Majoranas See arXiv:1201.5600
Bs(d)→µµµµ
A good example of flavour physics accessing high energy scales
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
17
CPV
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
18
Bs mass eigenstates: Weak eigenstates(mix via box diagram)
• q/p: complex number. |q/p| ≠ 1 CPV in mixing• complex amplitudes. ||≠ 1 CPV in decay
Even if not CPV in mixing or decay, you can generate CPV in the interference if
Main (but not only) experimental signature of a non-zero : it generates wiggles in the time-dependent angular distribution of the Bs→J/ψ →µµKK final state particles.The frequency of the (potential) wiggles is known: Δms.
Φs from Bs → J/ψ (µµ) KK
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
19
Φs from Bs → J/ψ (µµ) KK
wiggles …and this quantity is sensitive to BSM physics: LHT, non-MFV in SUSY-breaking lagrangian, ED..
SUSY-AC
Acta Phys.Polon.B41:657, 2 010
Warped ED with Custodial ProtectionarXiv:0812.3803
sin (𝝓𝐬 )
sin (𝝓𝐬 )
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
20
• Background: Events are weighted according to position in J/ψKK mass spectrum
• Angular distributions are distorted on data because of non-flat angular acceptance. Simulation (weighted according to kinematics seen on data) is used to correct for this
• Lifetime acceptance. Samples from different trigger lines are used to unfold trigger biases. Simulation is used for selection/reconstruction biases
Φs from Bs → J/ψ (µµ) KK
Analysis strategy: Fit the pdf of previous slide to data, considering experimental effects:
arXiv:1304.2600
arXiv:1304.2600
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
21
Analysis strategy: Fit the pdf of previous slide to data, considering experimental effects:
• Lifetime resolution: Non-perfect time resolution (45 fs, still much smaller than oscillation period, 350fs) convolved with the pdf. Main effect is a ~25% dilution of the amplitude of the wiggles. Measured on data using prompt J/ψ events
BJ/ψX
• Flavour tagging: The initial flavour of the Bs is determined either by a muon/kaon from the other B, and/or by a kaon from the fragmentation. The performance of these taggers is calibrated with control samples such as B+→J/ψK+, Bd→D*+µυ and Bs→Ds
- π+
Φs from Bs → J/ψ (µµ) KK
arXiv:1304.2600
22BSM after the first run of the LHC. Galileo Galilei Institute. Firenze,
2013
We perform the fit in bins of KK mass to better deal with non resonant component and, more important, to solve ambiguity of the equations
= 0.07±0.09±0.01 rad
Combined with Bs→J/ψππ
Φs = 0.01±0.07±0.01 radians
In good agreement with SM: -0.036±0.002(*)
(*)Penguins ignored
arXiv:1304.2600
sin (𝝓𝐬 )
Which, as in the case of for example Bs → µµ, sets constraints on BSM physics(Don’t get depresed by the plot, rememember comments in the Bs → µµ case)
Φs (results)
arXiv:1107.0266
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
23
Φs (ATLAS/CMS)
HFAG private/unofficial combination yields
ExperimentLumi. (fb-1) 4.9 5.0
ΔΓs (ps-1) 0.053±0.021±0.
0090.048±0.024±0.003
Φs 0.12±0.25±0.11
Set to 0CMS-PAS-BPH-11-006ATLAS-CONF-2013-039
≈ 0.00±0.07 rad
ATLAS and CMS also study Bs→J/ψ →µµKK , using 5 fb-1 eachBut only ATLAS reports a s measurement
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
24
B0→ K+p-
Bs→ K-p+
ARaw = -0.091 ± 0.006
ARaw = 0.28 ± 0.04
First observation of CPV in Bs decays
arXiv:1304.6173
CPV in B →Kπ cannot be calculated theoretically with accuracy, but combinations of observables allow building stringent SM tests such as:
LHCb measures the raw asymmetries (difference in observed yields between particle and antiparticle)
These are related to the CP asymmetries by being
Instrumental asymmetry production asymmetryattenuation dueto oscillation
±
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
25
First observation of CPV in Bs decays
Detection asymmetries is determined using D*+ → D0 (→Kπ)π. Value ~1%
Production asymmetry is obtained from the time dependency of the raw asymmetry
AP compatible with 0
Finally, we obtain:
Which, (for the moment) survives the Δ = 0 test
ACP(Bd→K+ π-) = -0.080±0.007(stat)±0.003(syst)
ACP(Bs→K- π+) = 0.27±0.04(stat)±0.01(syst)
arXiv:1304.6173
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
26
CPV in charm
2012- status
Vanishes if aCPind is 0 or if the time acceptance is
independent of the final state
We search for a direct CPV difference between D0→KK and D0→ππ
In SM it’s usually expected to be up to O(10-3), although recent works indicate it can be as large as several per mil.
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
27
CPV in charm (D* tag)
I) Count D0 decaying into charged Kaons and pions
II) Tag the flavour of the D0 at its production using events from the chain D*+ →D0π, seen as a peak in:
IV) Weight D0 phase space to cancel out experimental differences between kaon and pion samples
KK
KK
π π
π πIII) Measure asymmetries
LHCb-CONF-2013-003
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
28
CPV in charm (lepton tag & combination)
(3%)
Independent study using D0’s from semileptonic b→D0µX decays, where the D0 flavour is tagged by the accompanying muon of the D0 meson
• Similar strategy as for D*+ tags, incuding D0 phase space correction
• But different potential systematics/bkgs.
• In addition, existence of wrong tags (O(1%))
Analysis ΔACP (%)D*+ tag -0.34±0.15 (stat) ± 0.10 (syst)Muon tag +0.49±0.30 (stat) ± 0.14 (syst)Combined -0.15±0.16 (neglects <t>aCP
dir
term)
arXiv:1303.2614
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
29
Bd → K*(→Kπ) µµ
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
30
Bd → K*(→Kπ) µµ (LHCb analysis strategy)
• b→sµµ transition (like Bs → µµ)
• We select events using a BDT and special vetoes for specific backgrounds
• Correct (in an event-by event basis) for the effect of reconstruction/selection/trigger using simulation
• Validated on data via control channels (mainly Bd →J/ψ(μμ) K*(Kπ))
• Fit yields and angular distributions for observables in bins of q2 (dimuon invariant mass squared)
arXiv:1304.6325
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
31
Bd → K*(→Kπ) µµ (LHCb angular analysis)
arXiv:1304.6325
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
32
Bd → K*(→Kπ) µµ (LHCb angular analysis)
You can also reparameterize the fit pdf to get some cleaner observables:
arXiv:1304.6325
arXiv:1304.6325
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
33
Bd → K*(→Kπ) µµ
… And then you can compare to models
arXiv:1304.6325
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
34
Bd → K*(→Kπ) µµ
… And then you can compare to models
arXiv:1304.6325
Also CMS has results on this channel, see:CMS-PAS-BPH-11-009
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
35
Conclusions
• Flavour experimental data is a powerful test for BSM physics
• LHCb has plenty of results on beauty, charm and strange decays
• The BSM hint in charm CPV is vanishing
• Up to now, good agreement with SM. This allows constraining BSM parameter space
(in other words, we didn’t observe planet satellites or parallax…. yet )
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
36
Conclusions
• Most of our results used only 1 fb-1 (7 TeV).
• Publications with 3 fb-1 collected up to now are in preparation.
• The LHCb upgrade plans to collect 50 fb-1 at 14 TeV (equivalent to 100 fb-1 at 7 TeV)
• More precision (and new measurements) may finally show BSM (or keep constraining it)
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
37
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
38
Indirect approach
Eratosthenes
~2.3 k years till the direct observation…
• Low energy observables can access NP through new virtual particles entering in the loop indirect search
• Indirect approaches can access higher energy scales and see NP effects earlier:
• 3rd quark family inferred by Kobayashi and Maskawa (1973) to explain CP V in K mixing (1964). Directly observed in 1977 (b) and 1995 (t)
• Neutral Currents discovered in 1973, 10 years before observation of Z0
• Roundness of Earth (Eratosthenes, c.III B.C) discovered ~2300 years before direct observation
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
39
KS → µµ
arXiv:1209.4029
SM prediction :
Even if KL → µµ has been measured, KS → µµ remains interesting because it’s sensitive to different physics than KL → µµ(see arXiv:hep-ph/0311084)
In particular, if BSM is found in NA62, then limits/measurements of KS → µµ in the 10-11-10-12 range can be useful to understand its nature
LHCb (1fb-1) sets world best upper limit 9(11)x10-9 @90(95)%CLs
LHCb upgrade might be able to reach the 10-11-10-12 range thanks to improved trigger.
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
40
D0 → µµ
SM prediction: BR(D0→µµ) < 1.6x10-11
(Precision depends on knowledge of BR(D0→γγ) )
BSM physics (RPV, ED’s) can enhance it up to the 10-10 level
R-S warped EDarXiv:1212.4849
LHCb set an upper limit of 6.2(7.6)x10-9 @ 90(95) %CLs using 0.9 fb-1
Potential to reach more interesting region with LHCb upgrade
arXiv:1305.5059
Bs → µµ
Bd → µµ
D0→µµ KL → µµ
KS → µµ
3.2 <0.94 <7.6 6.84±0.11
11
LHCb LHCb LHCb BNL E871 LHCb
BR of neutral flavoured mesons into dimuons [x10-9]
BSM after the first run of the LHC. Galileo Galilei Institute. Firenze, 2013
41
Bs(d) → µµ
These decays are very supressed in SM
(note also the high TH precision)
BR(Bs → µµ) = (3.54 ± 0.30)x10-9
BR(Bd → µµ) = (1.07 ± 0.10)x10-10
(time averaged)Eur. Phys. J. C72 (2012) 2172, arXiv:1208.0934.
NUHM1 (SUSY) pre-LHC
arXiv:0907.5568v1
ED (R-S)arXiv:0912.1625
But several NP models could sizably modify those values, sometimes by orders of magnitude.
Whatever we measure, it impacts NP searches
top related